Abstract:
The aim of this thesis was the development of visualization techniques for cell therapy and interventional procedures with Magnetic Resonance Imaging. Two post-processing techniques for the detection and discrimination of magnetically labeled cells were successfully implemented and experimentally verified in homogeneous phantoms and in an ex-vivo bovine liver. Both techniques are based on the effects of magnetic field alterations in gradient echo sequences. Variations in the frequency domain were utilized to generate highlighted areas around the labeled cells in the image domain, while suppressing magnetically homogeneous areas. The techniques mainly differed in the application of a frequency domain filter and hereafter in the further post-processing steps in the image domain. The aims of both techniques were a short calculation time of the post-processing algorithm and detectability of small aggregations of labeled cells. The highlighting of the areas with frequency variations in the resulting image, in contrast to depiction of them in conventional imaging as hypointensities, is beneficial for discriminating magnetically labeled cells from natural low signal tissue. The developed techniques are helpful for the tracking of magnetically labeled cells in immune- or stem-cell therapy. Alterations of the magnetic field can also be utilized for the visualization of interventional devices. A basic spin echo imaging sequence was modified to release a trigger signal at an external output of the scanner. The trigger signal switched a direct current on a metallic conductor generating a magnetic field, which disturbed the homogeneity of the main magnetic field. A transient field alteration is generated, if the current is switched by the trigger only in certain time intervals during the sequence. In this case, distinct phase variations around the conductor, without image distortions in the remaining parts of the image, can be generated and assigned to the interventional device. The generated field alteration is controllable by amperage, thus the dimension of the phase variation in the image is scalable as well, which cannot be accomplished with visualization techniques based on susceptibility differences. This technique allows for flexible choice of sequence parameters to achieve the desired image contrast and simultaneously visualizes the interventional device without the need of additional scanning time.